Computer vision in bin picking
Smart Industry and IoT
Client company:Affix Engineering
Joep van Dijk
Jay Zievinger
Chaitanya Schoka
Deniss Ivanovskis
Project description
How can current technologies and methods of computer vision be optimized to improve the accuracy of bin-picking robots in detecting highly reflective, dark, and transparent objects under varying light contitions?
Context
The projects is in the domain of industrial automation.
The focus will be on improving robotic bin picking capabilities with a deeper focus on detecting, recognizing and picking reflective, dark and transparent objects. This is a challenge that is widely known in machine vision sectors with no proper solutions.
Results
The importance of deep learning within this problem, because of the difficulty with detection with the use of more traditional computer vision algorithms.
About the project group
For the past 5 months we've been working 4 days a week on this project together. We delved into the new topic of machine vision in the bin picking application.